Yes, the best combination for my program, Antigo, is to use (somewhat) more stochastic
moves for the first 5-9 ply. I decided to look into this after noticing how surprisingly badly my heavy playouts did as part of AMAF without UCT/tree search. A more stochastic version of my heavy playouts was much stronger when using just AMAF, but weaker when I used UCT. Switching after the first 7 plys gave me the strongest policy for AMAF with no tree. It also turned out to be the strongest policy when using UCT, although the improvement was small. - Dave Hillis -----Original Message----- From: Michael Williams <[EMAIL PROTECTED]> To: computer-go <computer-go@computer-go.org> Sent: Mon, 17 Nov 2008 12:01 am Subject: Re: [computer-go] FW: computer-go] Monte carlo play? It seems move selection in the playouts should be very random at first and more deterministic toward the end of the playout. Has anyone tried that?? ? Mark Boon wrote:? > > On 17-nov-08, at 02:42, George Dahl wrote:? > >> So you say that: "...I'm observing that most of the increase in level? >> comes from the selection during exploration and only in small part? >> from the selection during simulation.", could you elaborate at all?? >> This is very interesting. That almost suggests it might be fruitful? >> to use the patterns in the tree, but keep lighter playouts.? > > That's exactly what it's suggesting. But as I said, I need to do some > > > more testing to make a hard case for that.? > > Mark? > > _______________________________________________? > computer-go mailing list? > [EMAIL PROTECTED] > http://www.computer-go.org/mailman/listinfo/computer-go/? > ? _______________________________________________? computer-go mailing list? [EMAIL PROTECTED] http://www.computer-go.org/mailman/listinfo/computer-go/?
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
